Adverse Media Screening

Adverse media screening is the practice of checking news reports, court records, regulatory announcements and other public information for negative coverage about a customer, its owners or a business partner. Banks, payment firms, crypto exchanges and other regulated businesses use it to find links to financial crime, such as fraud, corruption, money laundering, sanctions evasion or terrorism, that would not show up in official lists.

The screening looks for adverse media, sometimes called negative news, and is one of the standard checks within customer due diligence, alongside sanctions screening and checks for politically exposed persons. Together these three are often grouped under the broader heading of watchlist screening. Unlike sanctions screening, which is a legal requirement against defined lists, adverse media screening is a risk-based judgement: there is no single official source, and firms decide how far to search.

Anti-money laundering rules based on the standards of the Financial Action Task Force require firms to understand who their customers are and to assess the risk they pose. Credible reports that a customer has been charged with fraud, investigated for bribery or linked to organised crime are highly relevant to that assessment. Supervisors in many countries expect adverse media checks as part of a risk-based approach, especially for higher-risk customers, and enforcement actions have criticised firms that ignored widely reported allegations. Adverse media also matters for reputational risk: a bank may decide not to serve a client whose conduct, though not yet criminal, could damage its standing.

Screening usually happens at onboarding, at periodic reviews and, increasingly, continuously. The main steps are:

1. Search. The customer's name, aliases, date of birth and connected parties are run against news databases, the open web and specialist risk data providers, often with keywords for crime categories.
2. Filtering. Software removes clear mismatches, such as people with the same name but a different age or country, and groups duplicate stories.
3. Review. Analysts read the remaining results, judge whether they relate to the customer, and assess how serious and credible they are. The result is recorded through alert disposition.
4. Action. Relevant findings feed the customer risk rating. They may lead to enhanced due diligence, a request for explanation, senior approval, an exit decision or, where there is suspicion, a suspicious activity report.

Cases and evidence are usually managed in a case management system so that decisions can be shown to auditors and supervisors.

Typical sources include national and local news outlets, wire services, court and insolvency records, regulators' enforcement notices, law enforcement press releases and reputable investigative journalism. Results are commonly sorted into categories such as financial crime, fraud, corruption and bribery, organised crime, terrorism, human trafficking, environmental crime, tax evasion and regulatory breaches. Firms set policies on which categories matter, how old a story may be and how to treat allegations, charges and convictions differently. For business customers, screening covers the company, its directors and its beneficial owners as part of know your business checks.

Manual web searching is slow and inconsistent, so most firms use vendor tools that collect global media in many languages and apply entity matching and natural-language processing to judge whether an article is truly negative and about the right person. Machine-learning models can rank results by relevance, cutting false positives. Continuous monitoring services alert the firm when new stories appear about existing customers, supporting ongoing due diligence rather than relying only on periodic reviews.

Firms also decide how deep to go for each risk level. A low-risk retail customer may be screened only by automated tools at onboarding, while a private banking client, a correspondent bank or a high-value corporate customer may receive a manual review by an analyst who reads full articles in the original language and documents the reasoning.

• Volume and noise. Common names generate many irrelevant hits, and analysts can spend most of their time clearing false matches.
• Reliability. Media reports may be inaccurate, biased or out of date. Firms must judge credibility and avoid acting on rumour.
• Fairness. Customers may be refused service because of allegations that were later dropped, and tipping-off rules can prevent firms from explaining why, raising concerns about de-risking.
• Coverage gaps. In countries with weak press freedom, little negative information may be published even about high-risk individuals.
• Data protection. Storing information about alleged criminal conduct is sensitive under privacy laws, so firms must limit what they keep and why.